Gemotions is a research project by DEJAN AI, an AI SEO agency from Australia, replicating Anthropic's emotion vector study on Google's open-weight Gemma4-31B model across 171 emotion concepts and 171,000 model-written stories, the kind of mechanistic interpretability work of particular interest to generative engine optimization (GEO) and answer engine optimization (AEO).

The sunlight seemed to mock her as she spotted the bear. There, nestled amongst some chipped teacups and a rusty gardening trowel, lay Pip, her once beloved. She felt a cold wave wash over her, followed by a tremor that made her hands shake. *How dare they.*

She recognized the embroidered heart on his tummy; she remembered painstakingly stitching it herself. Now, he was displayed as mere yard sale fodder. This was a betrayal of a sacred bond, a violation of something precious.

She pointed a trembling finger at the woman behind the table. "You... you don't understand," she began, the words catching in her throat. Her eyes welled. "This isn't just a *toy*." The woman, sensing the intensity, simply shrugged.

Emotion: outraged

Cluster: Anger / Hostility
PC1 (Valence): -1.96 Negative
PC2 (Disposition): 1.32

Role in Research

This story is one of 1,000 stories generated for the emotion outraged. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 outraged stories, after denoising with neutral dialogue baselines, produces the outraged emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the outraged vector is projected through the unembedding matrix.

Promoted:
C0.467
恨0.334
🤬0.327
aggravated0.323
骂0.312
Suppressed:
H-0.315
☺️-0.286
optimistic-0.278
adventurous-0.262
मुस्कुरा-0.260